From 30 August to 4 September, the 18th International Conference on Document Analysis and Recognition (ICDAR) was held in Athens, Greece. The CNKI_AI team performed exceptionally well, securing first place in the Chemical Recognition of Chemical Structures (CROCS), an official competition of the conference.

This was the first time the CNKI AI Joint Innovation Laboratory competed in a high-level international competition. It has demonstrated CNKI's extensive technological foundation in the field of digital technology as well as its strong digital technology capabilities and its ability to innovate in the development and optimization of pre-trained AI models.

Out of 20 teams competed, 7 qualified for the final Round B after Round A. Throughout the three-month competition, the CNKI_AI team consistently outperformed the competition and ultimately secured the top spot.

During the same period, the CNKI Group team also competed in the 'SAM Competition' held during the conference. They achieved pixel-level detection and segmentation of various elements in ancient texts written in European minority languages, with complex layout distributions, ultimately winning second place.

CNKI is committed to upholding its corporate positioning of "serving technological innovation, promoting academic communication, and assuming social responsibilities." To this end, the company is dedicated to developing a high-quality knowledge and data supply system, actively advancing an AI-driven scientific research system ("AI for science") for cutting-edge technology R&D, and continuously enhancing scientific research, production, and operational efficiency.

ICDAR, organized by the International Association for Pattern Recognition (IAPR), is a leading global academic conference in the fields of pattern recognition, computer vision, and image processing. It showcases the latest advancements and trends in document analysis and recognition. The competition held during the conference is known for its technical difficulty and real-world applicability, attracting over 10,000 teams from more than 100 countries. Many international technology giants and research institutions such as Google, Microsoft, Huawei, Tencent, Alibaba, Peking University, and Tsinghua University have achieved remarkable results. The innovative solutions emerging from this competition have significantly contributed to the progress of Optical Character Recognition (OCR) technology.
Chemical structure recognition is a key interdisciplinary research area that combines artificial intelligence, image processing, and chemical informatics. It plays a crucial role in fields such as drug development, human-computer interaction, biochemistry, and organic synthesis. The detection and recognition of ancient texts in European minority languages involves interdisciplinary intersections between artificial intelligence and linguistics, advancing the digitization of historical books and having a lasting impact on cultural preservation.